OutOfMemoryError when loading a ludicrously huge JSON animation on a Galaxy S3 #39
Comments
|
@subtleGradient Can you attach the AE file? There may be some things we can do in the meantime to reduce overhead. |
|
@subtleGradient so I spent way too many hours this week migrating over the JsonReader. However, the existing json structure of bodymovin makes it incredibly inconvenient unfortunately. I'm going work with Hernan from bodymovin to figure out a good solution here. |
Closed
gpeal
added a commit
that referenced
this issue
Jan 9, 2018
Previously, Lottie used JsonObject for deserialization. That meant that: 1) Deserialization is not guaranteed to be O(n) where n is the size of the json file. 2) The entire json string must be loaded into memory. Switching to JsonReader has the following advantages: 1) Reading is guaranteed to be O(n). 2) Large files can be read in buffers. However, deserialization code is much more cumbersome because you can't query for things like the existence of a key, the length of an array, and you can't re-walk the same part of the json multiple times. @felipecsl Did some work (#137, #139, #145, #152) a year ago to prepare to decouple parsing logic so that people could use jackson or some other method of deserialization. However, JsonReader is now the most optimal solution so some of the factory code can be simplified in a future PR. ## Performance Most animations deserialize ~50% faster than before. I was also able to deserialize a 50mb json file in 10s that couldn't come close to completing without OOMing before. |Animation|Old time (ms)|New time (ms)|% improvement| |----------|--------------|--------------|----------------| |Tadah|107|55|48%| |Nudge|100|51|49%| |Notifications|85|48|43%| |Star Wars|74|41|45%| |City|65|24|64%| |9squares|17|7|59%| |Empty State|9|4|56%| |Hello World|6|2|66%| |Hamburger Arrow|2|1|50%| Fixes #39
|
@subtleGradient Only took a year but I finally got this working :) |
Closed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
and others
JSON is 1.5 MB uncompressed, 94 KB zipped.
Lottie should probably switch from JSON to Flatbuffers. I'll open a separate issue for that.
The text was updated successfully, but these errors were encountered: